The objective of the present study is to define adiposity patterns in women. The study was conducted at a municipal health care center in Belo Horizonte, Minas Gerais, with 827 women aged 12-65 years. All participants performed several anthropometric measurements and answered a questionnaire with socio-economic, demographic, obstetric and life style variables. Factorial analysis was used as a multivariate technique to identify clustering and independence of anthropometric variables. Factor analysis identified three factors that explain 85.97% of the total variance. The first component was represented by skinfolds, the second by the Body Mass Index (BMI) and arm circumference (AC), and the third by the waist to hip ratio (WHR). These factors explained 38.42%, 29.03% and 18.52% of the variance, respectively. Positive associations were observed among all variables. Cluster associations were observed between thigh, chest, bicipital, sub-scapular and supra-iliac skinfolds (Factor 1) and AC and BMI (Factor 2) and WHR as a third factor. We may conclude that factorial analyses have adequately identified three adiposity patterns in this group of women: peripheral, global and central. The strong association between BMI and AC demonstrates that AC could substitute with advantage BMI in population screening for obesity. The diagnosis of obesity would be more precise if AC, WHR and thigh skinfolds were included simultaneously.
Obesity; Adiposity; Factorial analysis; Women; Anthropometry
